Research on AIDS and AIDS-related complexes (ARC) indicates that the disease is contracted by exposure to HTLV-III carried by an infected person. Numerous risk-factor studies also indicate that homosexually active men with multiple partners who engage in sex activities permitting HTLV-III exposure are at greatest risk for AIDS and ARC. The study of interventions which alter the behavior of high-risk individuals will lead to the development of prevention programs that reduce AIDS prevalence. The purpose of the proposed project is to evaluate the impact of an AIDS risk-behavior reduction program for homosexually active, at-risk, but apparently healthy males. Ninety subjects will be randomly assigned to either an experimental or waiting-list control condition. Following assessment on a range of self-monitoring, self-report and behavioral measures, experimental group subjects will receive a 12-session behavioral training intervention to reduce casual, high-risk sexual encounters. Training will include self-management (control) skills, assertion training to refuse high-risk propositions, risk reduction education, and skills training to promote stable rather than promiscuous relationships. Following treatment, experimental group subjects will be compared with controls on multiple measures of risk behavior, including frequency of casual sex encounters, type of sex activities, drug use, depression, anxiety, and health-related characteristics. The treatment will then be replicated with those subjects originally assigned to the control group and all subjects will be assessed at one-year follow-up to determine maintenance of behavior change. Contingent upon successful findings, a detailed training protocol manual will then be developed and distributed to other investigators and to public health agencies. This activity is intended to promote efficient, rapid dissemination of study procedures to other scientists, applied researchers, and health agencies.